Downloads: 0
Research Paper | Computer Science and Engineering | Volume 15 Issue 4, April 2026 | Pages: 207 - 211 | India
Integrating Digital Forensics and Machine Learning for Retail Return Fraud Detection: A Comprehensive Review
Abstract: Return fraud in retailing is the misuse of the return policy offered by stores for personal gain through the return of goods, which causes substantial losses to the retailers every year. Others include "wardrobing" which is returning heavily priced clothing after only wearing it once or twice, and returning stolen components, empty packaging, and even fake receipts. Indeed, return fraud causes retailers to hike even higher prices. The field of digital forensics provides an effective remedy by playing the role of a virtual detective that tracks down digital trails left by cases of fraudulent returns. Methods such as device fingerprinting, analysing metadata, and confirming shipping information are some of the ways that detect a pattern of fraud, including that of a fictitious identity and that of a device. Artificial Intelligence and Machine Learning technologies further amplify these functions by analysing massive data, recognizing anomalies in consumer patterns, and leveraging computer vision to verify the image of a product. The study emphasizes a shift from traditional physical checks to digital forensic strategies integrated with AI, addressing emerging fraud trends and reducing reverse logistics costs. By bridging gaps in current practices, this research provides actionable insights for academia and industry, supporting profit protection while maintaining customer satisfaction.
Keywords: Return Fraud, Digital Forensics, Artificial Intelligence, Retail, Wardrobing
How to Cite?: Girish Kurkure, Dipita Dhande, Manisha Shirsath, "Integrating Digital Forensics and Machine Learning for Retail Return Fraud Detection: A Comprehensive Review", Volume 15 Issue 4, April 2026, International Journal of Science and Research (IJSR), Pages: 207-211, https://www.ijsr.net/getabstract.php?paperid=SC26211110556, DOI: https://dx.dx.doi.org/10.21275/SC26211110556